Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Solenoid Electromagnet with strong core, low residual magnetism

  1. Jul 25, 2007 #1
    Hello, this is my first post here.

    I'm building an electromagnet as part of a larger experiment in my lab. My magnet needs to move a magnetic thread made with a paramagnetic substance. This string just barely responds to strong magnetic fields (tested it with standard speaker magnets). The electromagnet I'm building thus needs to generate a strong enough magnetic field to move this piece, even just so slightly. I can't afford to use a ferrite core because of hysteresis effects - I need to rapidly switch between multiple electromagnets and cant afford to have hysteresis causing a delay. Any suggestions on how I can do this? Suggestions on any core that I can use to amplify the field that has a really low coercivity? or any configuration (i'm currently trying a regular solenoid, maybe horseshoe solenoid?, dont know) or even any other method that I can employ to generate the field.

    To summarize, I need to switch between multiple electromagnets to move a magnetic material (that responds poorly, hence a strong field is required) and need a solution on magnifying the field intensity that doesn't involve expensive cores (mumetal costs a lot, i think).

    Big question, I'm sure. I'm using a Gauge 20 wire and have a standard power supply for power (about 3 amps max).

    Tried higher gauge wires - more turns, easier to wind - but they heat up too much and the field isn't impressive anyway.

    Appreciate any help in advance,

    Thanks!
     
  2. jcsd
  3. Jul 25, 2007 #2

    xez

    User Avatar

    Well there are lots of materials (different ferrites, et. al.)
    that are used in magnetics up into the MHz and GHz ranges
    of frequencies, so it is by no means impossible to have them
    respond quickly enough to follow fast changing signals
    despite their hysteresis.

    The bigger problem with fast changing fields is that the
    materials that respond more quickly (ferrites, powdered
    iron, et. al.) tend also to have much lower permeabilities.
    Ferrites, specifically, also tend to have relatively low
    saturation magnetic field strengths also. So though they
    can respond quickly, there's a fairly low limit as to the
    magnetic field strength they can generate.

    If you're looking for a magnetic field from several thousand
    gauss to perhaps in the near-1-tesla range, AND you need
    to change the field quickly, it's often preferable to just
    use good old air core electromagnets with high
    Amp-turn values. That way the hysteresis will be
    nearly zero, the inductance can be made manageably low,
    and there's no non-linear limit to the attainable field
    strength. Also the wire heating will be limited by the
    fact that you're seemingly operating in pulse mode
    where any given coil / electromagnet will be turned on
    then off quickly. In the place of a core use a water
    cooling setup to jacket and wind the coils around if
    further cooling is needed.

    Take a few big deep cycle marine / golf cart batteries,
    some relatively heavy gauge magnet wire (12g?), wind a
    few turns so that the resistance and inductance limits the
    current to something thermally managable (12G
    melts at ~ 235A in open air convection, so assume
    you're not going to sustain more than around 100A for
    long even with flowing cold water cooling), and see
    how that works for you.

    Use an appropriately rated IGBT for switching the currents,
    and use much heavier gague wire for the connections
    from the battery terminals to the coils -- like 4G or 0G
    or whatever .. moderately heavy automotive jumper
    cables are about right.

    You'd end up with something like 400 Amp-turns or more
    over a ring that could be an inch in diameter or less;
    that'd probably get the attention of your paramagnetic
    thread.

    For a test just get a car battery and a pair of jumper
    cables and wind a little coil out of 12G solid wire from
    some household ROMEX or something and use heat-shrink
    tubing or electrical tape to insulate it and manually
    use the jumper cables to generate a swiping contact
    of less than a second and see if it reacts well.

    I don't know how long your pulses are to be, or what your
    dimensional field requirements are, et. al. or I'd venture
    to give you more specific coil winding data.

    Also just because it's paramagnetic doesn't mean that
    that's the best way to mechanically couple to it to cause
    motion -- consider possible electrostatic effects depending
    on its dielectric properties, as well as things like using
    air bursts or mechanical vibrations et. al.
     
  4. Jul 25, 2007 #3
    Thanks for that XEZ,

    I understand what you're saying that I do. Unfortunately, the experimental settings require that the current injected be of a fairly low value, given that I'll be using a DAC board set to current output. Haven't specified which DAC board i'll be using, but i believe the current output is still in the 1 or 2 amps range for the DACs we have around here. And thats one of the problems, the current output has to be that low for applications like these and that compounds issues elsewhere. I'm using the 20G wire currently.

    The pulse length depend on how quickly i can get the thread to respond to my field. If it responds quickly, and the hysteresis is low, then that would mean the pulse can be short. Would a field be generated instantly for a 20G wire with low amp? Can I use some kind of Soft magnetic material? Soft ferrite maybe?

    Thanks
     
  5. Jul 25, 2007 #4

    xez

    User Avatar

    Well the rate at which the field establishes when you're
    using a magnetic core material and electromagnet will
    depend on a couple of things -- a) the core material's
    hysteresis curve and physical volume will dictate how
    much energy you have to apply to switch it in a given time.
    b) the inductance of your coil will cause the current
    to rise at a rate depending on your voltage.

    Inductance:
    1 Henry = 1 Volt * Second / Amp, so
    10V for 1 second would cause the current to ramp by 1A.
    1 Henry also equals one Weber / Amp.

    Magnetic flux density:
    1 Tesla = the strength of a very strong Neodymium
    rare earth magnet, probably many times stronger than
    the speaker magnet you're likely to have used.
    At 1 Tesla fields you're hard pressed to be able to pull
    a magnet off of an iron surface if it's even possible;
    it's uncontrollably strongly attracted within a few
    millimeters of an iron surface, and it'll accellerate so
    quickly together it'll either shatter or throw sparks.

    1 Tesla = 1 Weber / square meter

    B [in Tesla units] = u_r * u_0 * H

    Magnetic field intensity:
    H = Amps * Turns / meter


    u_r = relative permeability of your core as a ratio
    compared to free space.

    u_0 = permeability of free space = 4 * Pi * 10^-7

    So since your speaker magnet's insufficient, I assume
    you're talking about flux densities of around
    0.25 to 1 Tesla.

    Those kinds of fields are just about impossible to generate
    in a compactly sized soft electromagnet operating at
    3 Ampere current levels; you'd need a lot of turns and
    a moderately sizable core volume of quite high
    permeability and saturation field strength. It's very
    possible, but when you start using dozens to hundreds
    of turns and high permeability cores your inductance
    and hysteresis energy becomes such that rapid pulsing
    (e.g. several kilohertz) becomes non-trivial in power
    requirements and hence voltage and current levels.

    What are the rough physical dimensions you're able to use
    for a single pole-piece of the magnet, and then how
    much bigger and in what proximity can the rest of the
    electromagnet be? How close will the pole piece get
    to the sample, and what's the needed area of pole piece
    to become comparable to or larger than the area of the
    sample?

    Anyway I'll take a look at a few of my references to see
    if I can find materials that have saturation flux
    densities in the 0.25 to 1.0 Tesla range, and
    relative permeabilities in the 10^3...10^4...10^5 range.

    Certainly many of the classic ferromagnetic materials
    meet the saturation field and permeability requirements,
    but they're not generally appropriate for <= 100uS
    switching unless you apply a LOT of electrical power.

    Could you just use a Nd rare earth magnet and use
    the DAC to control a voice coil (e.g. speaker, galvanometer,
    or alternatively a piezo stack) to physically *MOVE* the
    small powerful rare earth magnet to modulate its
    field intensity at the thread's location at reasonably
    high frequencies? A little creativity with a stereo
    amplifier and speaker could get your mechanical
    magnet motion thing going.
     
  6. Jul 25, 2007 #5

    berkeman

    User Avatar

    Staff: Mentor

    Just a thought, but why use a magnetic field if the string just barely responds to a magnetic field. Why not move it some other way? Even if it has to be non-contact, just blast it with compressed air or something?
     
  7. Jul 25, 2007 #6
    Hey, thanks a bunch for that info XEZ,

    The sample that I mentioned has a diameter of around 40 microns but its about 30 cms in length - like a long thread. Its pretty light and I believe it's used for microfluidics applications - so thats the scale I'm talking about. It weighs less than a gram, and just one tip needs to be moved maybe 1 cm or so. The other electromagnets need to be close by (same distance) for the switching to move the sample this distance.

    The idea you suggested is creative. The ones that I've tried involve reconfiguring the whole setup physically to then induce some tiny motion magnetically. Is a Nd magnet available off-the-shelf? I understand that the requirements would demand stronger, more powerful magnets, but am hoping to contain it within a reasonable space.

    Berkeman, the functions of the electromagnet and the DAC would overlap with some other functions of the experiment as a whole :grumpy:. Turns out the sample needs to be moved magnetically.
     
    Last edited: Jul 25, 2007
  8. Jul 25, 2007 #7

    xez

    User Avatar

    You can get Nd magnets off the shelf; peruse the inventory
    of these places and you'll generally find several available
    varieties:
    http://www.sciplus.com/category.cfm?subsection=11&category=117
    http://www.goldmine-elec-products.com/products.asp?dept=1100
    http://www.alltronics.com/cgi-bin/category/34


    e.g.
    http://www.sciplus.com/category.cfm?subsection=11&category=117

    MICROMAGNET
    click here for color photo Mighty Micro Magnet tell a friend about this item so they can check it out
    Absurd strength in this rare earth magnet proves once again that size doesn't matter. Half the size of your pinky nail, the 3/16" x 3/8" x .05" thick magnets can lift 7 oz. of steel and are polarized on their ends not on their flat sides. It's .5 grams and strongest on its ends. Great for use with reed switches. Very, very bad to have around credit cards.
    on sale 32436 MICROMAGNET $1.50 / PKG(4)
    (was $2.00) add to cart


    NEO BAR MAGNET
    click here for color photo Itty Bitty Shiny Bar Magnet tell a friend about this item so they can check it out
    This neodymium bar is a slight 1/4" x 19/32" x 1/8" thick, but strong for its size. Little as it is, according to our strength and fitness lab, it can lift 3 lbs. Nice polished silver color, too.
    on sale 37408 NEO BAR MAGNET $2.95 / PKG(3)
    (was $3.75)


    http://www.sciplus.com/category.cfm?subsection=11&category=117

    Cylindrical Lift tell a friend about this item so they can check it out
    A good alternative for when you need to fit a magnet into a little hole, like in a door closure. These neodymium magnets are 13/32" dia x 19/32" long, with approx 4 lbs of lift. Cute and useful. Sold in packages of (2) or for larger savings try the package of (20).
    37403P2 4-LB CYLINDER MAGNET $3.95/PKG(2) add to cart
    37403P20 4-LB CYLINDER MAGNET $19.50/PKG(20) add to cart
    Cylindrical Lift
    click here for color photo

    http://www.sciplus.com/category.cfm?subsection=11&category=117


    click here for color photo Powerful Flat Disc tell a friend about this item so they can check it out
    Proportionally, this little neodymium magnet is way, way, way stronger than you. Stronger than you and your favorite football team, to tell the truth, since this little guy is only 1-1/2" dia x 1/4" thick, and he has a 25-lb pull. With a 3/16" hole in his center!
    37404 NEO DISK MAGNET $5.95 / EACH add to cart
     
  9. Jul 25, 2007 #8
    Hey thanks a bunch XEZ!

    This is great stuff! Lets see what I can do with these. Should get my string's attention finally. I'll check back with some more questions in a couple of days.

    Great job, Thanks again!
     
  10. Jul 25, 2007 #9

    xez

    User Avatar

    Since your thread has a small diameter, you'll want to
    use a magnet (or pole piece) that is sufficiently small
    or geometrically tapered / pointed in height so that the
    field is very highly spatially non-uniform in the vicinity of
    the thread and pole piece; that way the attraction force
    will be strongest due to the change in field energy vs.
    change of thread position.

    I think using some kind of mechanical actuator and
    an appropriate high strength Nd magnet moving in
    proximity to the sample may be most effective in
    field strength variance while being more compact in size
    than an electromagnet.

    However on the topic of electromagnets, it looks like there
    are various high saturation field ferrites
    (e.g. various Manganese Zinc) that have
    saturation field densities of between 0.4 and 0.58 Tesla
    or so.

    There is another material called 'MPP' Molypermalloy
    Mo-Ni-Fe which saturates around 0.75T.

    There's another material, Kool Mu which saturates
    around 1.05T.

    There's another material called 'High Flux' that saturates
    around 1.5T.

    The 'High Flux' may be your best bet due to high
    saturation flux density as well as permeability,
    remanence, et. al. though MPP wouldn't be a bad
    choice, and KoolMu probably better than most ferrites.

    You'd have to get either an 'E' core or cylinder or
    half of a pot core, or saw/break a toroid for an
    electromagnet. Wind it so that at your nominal coil
    current the Amp-Turns gives enough Oersteds of (H)
    in the core that H * mu will give you a nearly
    saturated flux density and that's the best that you're
    going to do with that size/material of core.

    You'll still need to supply a fairly substantial amount of
    energy quickly to rapidly switch the magnetization
    of the cores, though, all the way from saturation on
    one side of their curve into saturation on the other side.
    The benefit of using the right material is that you get
    high field strengths at at least a somewhat beneficial
    permeability factor by having the core there versus
    having just air.


    http://www.coilws.com/Powder Core Inductor/PowderCore_Guide.php

    http://www.mag-inc.com/powder/2006_Magnetics_Powder_Core_Catalog.pdf

    http://www.cwsbytemark.com/CatalogSheets/Ferrite_Materials[1].pdf

    http://www.cwsbytemark.com/CatalogSheets/spec_sheets.php

    http://www.cwsbytemark.com/CatalogSheets/MPP PDF files/6.pdf

    et. al.

    (Molypermalloy Mo-Ni-Fe),
     
  11. Jul 25, 2007 #10
    Ok XEZ,

    That one got some neurons seriously fired up! :D Going to take me a while to peruse the catalogs. I'll do that and get this working soon. Will inform you on how it turns out.

    Thanks a bunch again!!!

    Would give you some stars if i could.

    See you in a bit.
     
  12. Jul 26, 2007 #11

    xez

    User Avatar

    You're quite welcome; good luck!

    Just keep in mind that when you work with strong
    rare earth magnets, you must assume that you're going
    to have zero physical control of them anywhere within a
    couple of inches of a ferromagnetic substance or another
    magnet.

    They're ceramic, so they will chip / shatter easily if
    they're allowed to have things clunk against them.

    You're likely going to be unable to peel them off direct
    contact with a ferromagnet or other magnet without
    massive difficulty.

    So usually it's best to keep them well away from such
    objects, and when you're intentionally going to bring
    them close do so only if they're physically distanced by
    at least 0.1" or so of intervening strong plastic/wooden
    material.

    To make some kind of apparatus where they are intended
    to be close to another magnet or ferromagnetic material,
    you'd devise some kind of mounting mechanism for the
    individual magnet where it's clamped in place firmly but
    gently (because they're shatter like glass is the pressure
    or abrasion is too high in any spot). Then once the
    individual magnet is clamped to a holder, you can use some
    mechanism like a screw / vise / whatever to adjust
    the distance from the magnet holder to get the close
    proximity you need while being able to control / adjust
    the spacing and prevent things from launching themselves
    together.

    You'll get the best field concentration with a design
    where a north pole is very close to a south pole and in
    between those is your field area. Join the opposite
    'unused' pole ends via a conduit of iron if possible
    so there'll be a high permeability flux path in a fairly
    complete circle with the primary air-gap being your
    sample area.

    It may work ok without a nearly closed high permeability
    magnetic path, and with using only one magnet instead
    of two facing pole pieces, but if you need more field
    intensity in the area, using the gap between
    paired pole pieces, is always an option to consider.
     
  13. Jul 26, 2007 #12
    Hey XEZ,

    I was thinking - are the cores that you mentioned - MPP/High Flux/Sendust available only as circular cores? Because that would mean that the field lines would be completely within the core itself right? Would it stick to an iron surface if the wire is wound in a toroidal fashion? Are these available as straight rods, instead of just circular cores? I think you understand what I'm getting to - I'd need a field that comes out of the solenoid fairly significantly. And how hard is it to break a large MPP core, so that I get a horseshoe type core which I can then use?


    The price for an MPP 0.380 inch O.D. powdered core is $0.80. Do they really have such low hysteresis effects? I imagined that for high permeability (low amp-turns), low hysteresis material, the price would be really high, unless maybe this is no big deal. Can I use this ( http://www.cwsbytemark.com/prices/mpp.php ) for my switching function without the residual effects?

    At the moment, I believe this magnet would serve my purposes since it is small enough that for a low amp - say 1 A, I'd get a pretty decent magnetic field : 0.05 T maybe for around 23 turns for 20G. The sample and the magnet would be separated by some medium, maybe wood or acrylic.

    Am I missing something? Somehow sounds too good to be true - maybe the fact that the field stays mostly within for a toroidal winding. If i got a thicker core and wind the wires around the circumference, what would the direction of the field be given a hollow center?

    Is there any way that I can measure how 'far' the magnetic field would actually be strong, i mean - theoretically. For how much distance is this Flux density maintained?

    Also, there are a couple of terms that I don't understand from the catalogs:

    1. In the specs for these cores, they mention Turns and Single Layer turns - whats the difference? and which should I consider?
    2. Whats "u.w.f" ?
    3. Whats the Mean magnetic Path Length?
    4. Is there some page listing the hysteresis loss coefficient for these material so that I can calculate the Core loss, along with the other losses?

    Also, passing this amount of current for a winding this thick - the resistance being low, wouldn't that heat the winding up? Say, for a longer pulse.

    Thanks again,

    This experiment just got more interesting.
     
    Last edited: Jul 26, 2007
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Have something to add?



Similar Discussions: Solenoid Electromagnet with strong core, low residual magnetism
  1. Electromagnet cores (Replies: 9)

Loading...